Which Track Record Window Should You Trust: 7, 30, 90, or All-Time?

How long should a trading track record be before you trust it? Long enough to have survived at least one bad stretch, and backed by enough closed positions that the numbers are not noise. There is no magic day count, but a seven-day number with three closed trades and a three-year number with four hundred are not the same kind of evidence, and the length of the window is the first thing that tells them apart.

PickVesting computes every record from a read-only brokerage connection and shows the same metrics over four windows: 7 days, 30 days, 90 days, and all time. The point of showing all four is not to let you pick the flattering one. It is to let you compare them, because the gaps between windows often say more than any single figure.

Why a short window flatters almost everyone

A short window is easy to win by luck, so a seven-day return should move you the least. Over seven days a trader can post a large number from one or two positions that happened to go the right way. That is not skill you can lean on; it is a sample too small to separate a good process from a good week.

This is why on PickVesting a figure never appears on its own. Every return is shown with how long the record has been verified and how many closed positions it covers. A 40% seven-day return next to "2 closed positions" is a different sentence than a 40% return over three years across hundreds of positions, and the metadata is right there so you do not have to guess.

Use the short windows for one thing: recency. The 7-day and 30-day numbers tell you whether the account is active and how it is trading right now, not whether the trader is worth following.

What 90 days starts to show

Ninety days is the first window where a style has to show up more than once. A quarter usually contains a mix of conditions, so the 90-day return, win rate, and max drawdown begin to reflect how a trader behaves across more than a single setup.

Pay attention to max drawdown here. A 90-day return that looks smooth with a small drawdown is telling you something different from the same return with a deep drawdown in the middle. Both can end at the same place, but they describe very different experiences of holding those positions. For how each of these figures is defined and computed, see How PickVesting Calculates Return, Win Rate, and Max Drawdown.

Ninety days is still not long. It is enough to disqualify a record, less enough to confirm one.

Why all-time is the window that matters most

The all-time window is the one to anchor on, because it is the only one that can show whether a trader survived a stretch they did not choose. Returns computed over a longer verified history have had the chance to include drawdowns, losing runs, and the trades nobody screenshots.

Two things make the PickVesting all-time number harder to dress up:

• The record covers the whole linked account, every position in stocks, options, and crypto, including contracts that expired worthless. Losers are not quietly left out.
• Snapshots are append-only. A correction is a new row, and nothing is rewritten, so the history you read today is the history that was recorded on each of those nights.

All-time is also where the closed-position count carries weight. A long record with few trades may just be a buy-and-hold account; a long record with a high position count is a trader who has repeatedly made decisions you can inspect.

Read the windows against each other

The most useful move is not to trust one window but to compare them, because the contrast is where the signal is. Here is a rough guide to what each window answers.

Some comparisons to run in your head:

1. If the all-time number is strong but the 90-day number has collapsed, recent trading does not match the lifetime claim. That is not disqualifying, but it is worth understanding.
2. If the 7-day number is spectacular and the all-time number is modest, you are looking at a hot week, not a track record.
3. If every window is roughly consistent and the position count is high, you have the most coherent kind of record.

Check length and linkage together, not just length

A long window only means something if the record is actually attached to everything, so read the window alongside how the account is linked. A three-year record on one of three accounts is a three-year record of one slice.

PickVesting makes this visible in a few ways:

• Every profile states how many brokerage accounts are linked, so an unlinked account is a visible absence rather than a secret.
• A trader can disconnect, but the record keeps everything up to that day and shows the gap after it. Disconnecting never erases history, so a long record with a recent gap reads as exactly that.
• The connection is read-only through SnapTrade, across brokers like Robinhood, Charles Schwab, Fidelity, Interactive Brokers, Webull, and more, read-only via SnapTrade. It can read history and holdings and nothing else.

So "how long" is really two questions: how long has it been verified, and over that time, how much of the trader's activity was inside the linked accounts. The second question is why length alone is not enough. You can browse verified records on Discover and see the window figures, position counts, and linked-account counts side by side.

A practical rule of thumb

Treat the windows as a sequence of filters rather than a menu. Start at the top and work down.

1. Read the all-time return next to its verified length and closed-position count. Few positions or a very short history? Stop treating the number as evidence.
2. Read all-time max drawdown. This is the stretch you would have had to hold through.
3. Read the 90-day window. Does current trading look like the lifetime record, or like a different trader?
4. Glance at 7-day and 30-day only for recency and activity.
5. Check linked-account count and any gap in the history.

None of this tells you to follow anyone. It tells you whether the record is worth the time it takes to read it.

What this still does not prove

A long, position-rich, fully linked record is strong evidence about the past and nothing more. A record proves what happened in the linked accounts and says nothing about accounts that are not linked. It does not know your suitability, your position size, or your taxes, and it does not predict the next trade or future results.

It also does not vouch for advice in a chat room, and your fill price will differ from the creator's. PickVesting does not place trades for anyone and does not offer copy trading. A record that survived three years is a reason to read carefully, not a reason to copy blindly. The disclosures spell out the limits in full.

If you want to see how these windows read on real accounts, start on Discover and compare the figures against their position counts and linked-account counts.

PickVesting is not an investment adviser and does not provide personalized investment advice. Past performance is not a prediction of future results.